The experimenter did not perform the pain simulation at this age, but rather a sadness simulation (we thought that children at this age might be suspicious if two similar pain simulations were presented to them). The experimenter told the child excitedly that she brought her favorite doll (unisex doll of a cartoon figure) but then “discovered” that the doll’s arm had been broken. She feigned sadness for 50 s, without making eye contact, alternating between holding the doll (first 30 s), trying to fix it, and placing it between her and the child (remaining 20 s). Finally, the experimenter succeeded in fixing the doll and was happy. If the child was able to fix the doll at any point, the simulation ended. Similar simulations have been used to measure young children’s empathy and prosociality (e.g., Dunfield and Kuhlmeier, 2013 (link)). Prosocial behavior in this task was coded dichotomously (0 = not shown, 1 = shown), as well as on a 4-point scale reflecting the extent of assistance shown by the child: 0 = none, 1 = brief (a single or weak attempt), 2 = moderate (child tried to help/comfort a few times, or made a single intense or complex attempt), 3 = prolonged (child repeatedly and substantially engaged in prosociality). A 3-point spontaneity score was also coded, reflecting whether the child tried to fix the doll spontaneously, that is, even before the experimenter demonstrated how it might be repaired by trying to fix it herself, with 0 = no prosocial behavior, 1 = acted prosocially, but not spontaneously, 2 = spontaneous prosocial action. Inter-rater reliabilities, based on 20% of the sample, ranged from ICC = 0.97 to 1.00 for all the codes.
Free full text: Click here
